The present invention relates to a receiving technique, and more specifically, to a technique of diversity-receiving a transmission signal of Orthogonal Frequency Division Multiplexing (OFDM) system, for example.
The OFDM system is one of multi-carrier transmission systems that divides a transmission signal into a plurality of carrier waves to transmit the signal, and is used in various fields such as digital television broadcasting.
Diversity receiving is performed in mobile terminals or in-vehicle receiving apparatuses employing the OFDM system in order to improve a reception quality.
A receiving apparatus that diversity-receives a transmission signal of OFDM system performs Fourier transform on a reception signal of each branch to obtain 52 sub-carrier signals corresponding to 52 sub-carriers from “−26” to “+26” having different frequencies, as shown in FIG. 15, for each branch for one symbol. A sub-carrier signal of each branch is combined for each sub-carrier by the algorithm called Maximum Ratio Combining (MRC) to obtain a combined signal. Before Fourier transform, automatic frequency control (AFC) that corrects a carrier frequency error is also performed on the reception signal.
Further, Fast Fourier Transform (FFT) is known as an algorithm that performs Fourier transform in a high speed, and FFT processing is typically performed in the receiving apparatus as the Fourier transform.
Techniques from various viewpoints have been proposed for such a receiving apparatus.
For example, Japanese Unexamined Patent Application Publication No. 2006-101245 discloses a technique that mitigates the effects given by variations of a frequency error estimation value for each branch on correction of frequencies. According to this technique, a signal intensity and a frequency error of a signal of each branch are detected in AFC performed on the reception signal before FFT processing. The frequency error of the reception signal of each branch is weighted and combined according to the signal intensity, and a phase of each branch is corrected based on the combined frequency error.
Further, Japanese Unexamined Patent Application Publication Nos. 2006-80624, 2006-253915, and 2010-226233 disclose techniques for improving a quality of a combined signal by modifying a combining method by MRC. Before describing these techniques, the combining method by MRC will be described first.
In the combining method by MRC, a transmission path response (also referred to as a transmission path power or a transfer function) for each sub-carrier of each branch is estimated, and a weighting coefficient to a sub-carrier of each branch is determined based on the transmission path responses that are estimated. A sub-carrier signal of each branch that is weighted is combined for each sub-carrier. When the k-th sub-carrier signal of the i-th symbol of the m-th branch that is subjected to FFT processing is denoted by Sm(i,k), for example, a weighting coefficient Wm(i,k) is calculated for each Sm(i,k), and combining is performed using a result obtained by multiplying each Sm(i,k) by the weighting coefficient Wm(i,k) corresponding to this Sm(i,k).
The weighting coefficient Wm(i,k) is expressed by expression (1). In the expression (1), Hm(i,k) denotes a transmission path response (transfer function) of the k-th sub-carrier of the i-th symbol of the m-th branch, and Hm*(i,k) denotes the complex conjugation thereof. Further, N denotes the total number of branches.
                                          W            m                    ⁡                      (                          i              ,              k                        )                          =                                            H              m              *                        ⁡                          (                              i                ,                k                            )                                                          ∑                              j                =                1                            N                        ⁢                                                                            Hj                  ⁡                                      (                                          i                      ,                      k                                        )                                                                              2                                                          (        1        )            
When a combined signal of the k-th sub-carrier of the i-th symbol is denoted by D(i,k), the combined signal D(i,k) can be expressed by expression (2).
                                                                        D                ⁡                                  (                                      i                    ,                    k                                    )                                            =                            ⁢                                                ∑                                      m                    =                    1                                    N                                ⁢                                                                            W                      m                                        ⁡                                          (                                              i                        ,                        k                                            )                                                        ×                                                            S                      m                                        ⁡                                          (                                              i                        ,                        k                                            )                                                                                                                                              =                            ⁢                                                                    ∑                                          m                      =                      1                                        N                                    ⁢                                                                                    H                        m                        *                                            ⁡                                              (                                                  i                          ,                          k                                                )                                                              ×                                                                  S                        m                                            ⁡                                              (                                                  i                          ,                          k                                                )                                                                                                                                  ∑                                          j                      =                      1                                        N                                    ⁢                                                                                                                                    H                          j                                                ⁡                                                  (                                                      i                            ,                            k                                                    )                                                                                                            2                                                                                                          (        2        )            
As shown in the expression (1), the weighting coefficient Wm(i,k) depends on the amplitude of the transmission path response (transfer function) H. Specifically, in the combining method by MRC, a large weighting coefficient is multiplied by a signal with large amplitude of the transmission path response H, and the signal of the branch is enhanced. However, even when the amplitude of the transmission path response H is large, it does not necessarily mean that the signal of the branch is strong or the C/N ratio of the branch is good.
For example, while a signal of one branch has a poor C/N ratio and small amplitude, the amplitude may increase due to an operation of auto gain control (AGC) and the amplitude of the transmission path response H may increase. When combining is performed by the combining method by MRC in such a case, the C/N ratio of the combined signal becomes poorer than the C/N ratio of the signal of the branch with good C/N ratio due to an influence of a signal of a branch with poor C/N ratio.
The techniques disclosed in Japanese Unexamined Patent Application Publication Nos. 2006-80624 and 2006-253915 calculate a carrier to noise ratio (C/N ratio) of a symbol signal of each branch after FFT processing or a weighted value (first weighted value in Japanese Unexamined Patent Application Publication Nos. 2006-80624 and 2006-253915) based on the relative ratio of these C/N ratios, and multiply the symbol signal of the branch and a transmission path response H estimated for the branch by the weighted value, and then perform combining by MRC.
The technique disclosed in Japanese Unexamined Patent Application Publication No. 2010-226233 determines a level of an interference wave by calculating a modulation error ratio (MER) after FFT processing, weights the signal of each sub-carrier by a weighted value according to the determined level, and then performs MRC combining.